Monday, 25 March 2019


Guest Post by PlanetNL
PlanetNL23: Backup Power Risks Reveal More Pre-Sanction Shenanigans

Two new reports on the Public Utilities Board (PUB) website shatter the illusion that the Muskrat Falls project will avoid the need for oil-fired generation in the province.  Upon decommissioning of the 490MW steam generation capacity at the Holyrood Thermal Generating Station, there is a critical reliability scenario arising in the all too likely event the Labrador Island Link (LIL) experiences an outage that will lead to a 500MW capacity deficit on the Avalon Peninsula.  Another report identifies that even when Muskrat runs optimally, there is insufficient capacity reserve without the addition of two new 58.5MW Combustion Turbines (CTs).

This new information gives further evidence of how Nalcor distorted the comparison of the Isolated vs Interconnected pre-sanction scenarios to favour Muskrat.  As reality sets in, Nalcor finds itself left with only one viable solution: the installation of several new diesel CTs, likely on the very same Holyrood site.  The irony of building the colossal Muskrat Falls megaproject for the purpose of closing Holyrood, only to find that a direct replacement must be constructed, is an astonishingly shameful exposure of Nalcor’s deceit.

Report: Reliability and Resource Adequacy Study
This 620-page study includes a section explaining how the complexity and scale of the LIL poses considerable certainty of outage events that may last from just a few hours to potentially weeks or even months.  Unacceptable reliability is bad enough, but an extra plot twist comes as Nalcor provides just enough information to read between the lines that the off-Avalon regions of the Island ought to be adequately served by Island generation at all times.  The Avalon region, however, will be considerably shortchanged and condemned to suffer rotating outages.  
Further, Nalcor admits in this report they cannot assure any firm energy could be imported over the Maritime Link during winter and even if it could, they further admit transmission constraints won’t get the extra power to the Avalon anyway.  The choice Nalcor deliberately leaves unanswered, perhaps because the saga before the PUB has been mandated to last all year, is whether to endure many repeats of DarkNL-like rotating outages, or to meet best practice reliability criteria and build a substantial new fuel-fired backup system.

Number Crunching an LIL Outage
The Nalcor report includes the graphic below illustrating the energy shortfall across the Island on a cold winter day without the Labrador Island Link in service.    The chart clearly shows a minimum power shortfall of near 200MW rising to 500MW during the daily peaks.  Nearly one-third of the Island load would be unserved during daytime hours.   As the chart shows a pattern of normal power delivery though, it doesn’t accurately convey the magnitude of the problem: the required cold load pickup after an outage would feature much larger and broader peaks.  The bottom line is a lot of customers won’t get power without the LIL functioning at a high level.

By showing the entire Island demand, Nalcor deliberately veils the fact that the problem is not evenly distributed.  As all of the Island’s firm generation is off the Avalon, virtually all of those areas have the potential for adequate power supply.  Notwithstanding possible violation of reserve margins, the lights are likely to be on everywhere but the Avalon.  Nalcor executives are surely a bit shy to show the public a similar chart for the Avalon Peninsula sub-system only.  If they did, it would show nearly half the required demand cannot be served in the best case.  With cold load issues, the problem will only be worse. 

The crux of the issue is the Avalon Peninsula is connected to Island generation only by the Bay D’Espoir East transmission system having a limit of 650MW.  Avalon demand exceeds 1100MW on cold days thanks to the prolific use of electric heating.  Perhaps it’s no secret but it bears repeating: Holyrood runs every winter primarily, if not solely, to serve the Avalon region.

Does Nalcor think We’re Stunned?
At this time Nalcor is merely acknowledging the risk that will develop after closing the Holyrood power station.  They appear to sheepishly suggest that an Avalon backup system would be a good idea without admitting it is an essential solution.  Despite the absurdity, does Nalcor wish us to believe this was unforeseen?  This had to be completely foreseeable and it should have been included in the Muskrat Interconnected pre-sanction cost estimates as a direct project cost.  Why wasn’t it included?  Perhaps because the significant costs would have hurt their financial analysis and especially because a sacred objective sold to the public was to eliminate oil burning at Holyrood.  Had they said Holyrood would be replaced with another fuel burning power station of matching capacity even after building Muskrat, their project would’ve been tough, perhaps impossible, to get the public’s buy-in. 

Quantifying the CT Requirements
NL Hydro installed a 123MW combustion turbine in 2014 principally to mitigate the frequent loss of a unit at Holyrood (which performed exceptionally well this winter, by the way).  This CT would be maintained post-Holyrood, however, the new Nalcor report fails to provide an overall operating strategy.  The report simply presents the cost of 1, 2, or 4 new CTs of 66MW each.  It also provides an option for a 170MW Combined Combustion Turbine plant.  

Of those options, the correct solution to consider is a set of 4 x 66 MW CTs.  The CCCT is only advantageous for long-term steady baseload usage; it is entirely unsuited to act as an emergency standby unit.  It’s unclear of course whether the inclusion of the CCCT option means Nalcor may be implying we need a new baseload diesel-fired CT as part of the backup strategy – this would be an outrageous travesty if it is so.

Presuming this is an emergency backup system only, the question arises whether the addition of 4 x 66MW units would suffice.  Adding that to the existing 123MW unit totals 387MW.  Why not meet the full 500MW deficit Nalcor describes?

Nalcor does not acknowledge how they developed these alternatives.  Do they absurdly think LIL outage events won’t coincide with peak winter demand?  Logic suggests that winter peak demands will be the stressful event that breaks the system.  It therefore seems 6 units of 66MW each are needed to assure power reliability on the Avalon.

Perhaps Nalcor is leaving room for Demand Side Management to reduce peak loads.  Nalcor does acknowledge they and Newfoundland Power have just started a new DSM study.  Let’s hope they “see the light” this time and not just the LED-type.   Their previous joint studies appeared to be pre-determined from the outset to support the development of Muskrat and resembled nothing like that completed in other provinces.  We always hope the next time is different, but no one should hold their breath.  We are not that stunned.

The Cost of New Backup CTs
Nalcor estimates a capital cost of $665M for the 4 new CTs.  If we expand this to 6 units, the total becomes $1.0B.  To most of us, this is a clear Muskrat project cost (again Nalcor, we’re not stunned) but Nalcor is almost certain to say it’s an unrelated capital cost and that it will be a part of NL Hydro operations and rate base.  The 123MW unit should also be categorized as a post-Holyrood Muskrat requirement and part of the project cost.

The report also identifies significant O&M costs (other than fuel) that when extrapolated for 6 new units plus the existing 123MW unit will come very close to the O&M cost at the old Holyrood power station.  

Perhaps later this year, as PUB hearing participants push for an answer, Nalcor will explain their backup strategy better and a revised cost and rate impact estimate will be prepared.  For 6 new units, that rate impact will be in the ballpark of 1.5 c/kWh.  This of course is a new increase on top of the Muskrat rate increases the public is already anticipating.

If Muskrat generation and the LIL both run really well, there might be $100-150M/yr in fuel savings achieved.  Unfortunately, most of that savings will be absorbed by the financing costs of the new CTs.  As a result, there may be only about $50M/yr in net cost reduction achieved at Holyrood.  If Muskrat infeed via the LIL has a bad year, that will quickly go up in smoke.

Another Report: Marginal Cost Study Update
This consultant report is a tough read and based on plenty of tenuous and arguably creative assumptions to be considered in rate design, potential time-of-use rates and other measures.  Buried among this were a few key surprise details that intersect with the backup power CT issue.

Firstly, the report, covering the period 2019-2029, acknowledges there is virtually zero capacity and energy load growth across that time.  Later comes a statement, “Hydro’s planned capacity additions for years 2019-2029 include two 58.5MW single cycle combustion turbines.”  This is the first direct reference found in any Nalcor document that the revised load growth scenario requires such new capacity additions.  Further the consultant clarifies, “the Island Interconnected System is expected to be just capacity sufficient (14% reserves) in generation for years 2021-2029” which indicates the CT additions are vitally important to meet winter peak capacity needs.

This new finding that under no load growth, Muskrat is insufficient to replace Holyrood seemed markedly at odds with Nalcor’s pre-sanction plans.  Indeed, doublechecking the DG3 capacity plan validated by Manitoba Hydro International in 2012, confirms that 50MW of new CT capacity was needed in 2015 and then another 50MW in 2032 was the first post-Muskrat addition.  These additions were driven by the DG3 fantasy load growth scenario that predicted about 25% load growth by 2032.

Pre-Sanction Muskrat Estimates for CT Requirements Woefully Underestimated
Most people would think that if 100MW of new CTs met high-load conditions after Holyrood decommissioning, then none would be needed under today’s flat load requirements.  Apparently, this is not so. In fact, we need quite a bit more than 100MW despite the substantial decrease in the load forecast.  Not only is 117MW of new CT capacity going to be required but this is on top of the 123MW unit installed in 2014 that was never included in the DG3 growth forecast.  These add up to 240MW necessary to satisfy no load growth conditions.

Assuming today’s Nalcor is correct, then the old Nalcor crowd (which includes some of the same analysts) had their pre-sanction modelling severely wrong.  For a start they needed 240MW of CTs just to keep zero load growth reliability.  Then to meet their load forecast, they should have increased this to about 600MW.  An additional 400MW of emergency backup would also be needed (mostly if not entirely on the Avalon) to mitigate LIL outage risks as Avalon load growth continued to 2029.  It seems the DG3 plan was missing a massive $2B series of CT installations prior to 2029.  O&M and fuel costs through that period would also be substantial.   

The DG3 estimates upon which final sanction in December 2012 relied had indicated a $2.4B Cumulative Present Worth advantage for the Muskrat Interconnected Option.  That analysis, shown to be wrong in many areas, now appears to be exposed as having another major flaw.  Had the pre-sanction estimates included the required 1000MW armada of CTs – and more still in later decades – this one change alone would likely have reduced the CPW advantage to zero or worse.

Links to Reports: